We live in an epoch characterized by high levels of socio-economic inequality. The income Gini Index is a commonly used statistical tool for gauging economic inequality in an economy. A simple way of interpreting the index is as such: an economy in which every individual has the same income would have an income Gini Index of 0, while an economy in which a single resident earned all the income and everyone else earned nothing would have an income Gini Index of 1. The index thus ranges from 0 to 1, with higher values corresponding to higher levels of economic inequality. The US’ income Gini Index has steadily rose since 1970, reflecting the widening income equality plaguing the nation.

Data from the Pew Research Centre on median incomes and the share of aggregate US income by income tier further highlight the growing income disparities in the country.

This stark economic rift was brought to the very forefront of American politics in 2016 when Donald Trump won the presidential election - the 2,584 counties that Trump won contributed to a mere 36% of the country’s economic output, while the 472 counties Clinton won contributed to almost two thirds of the US economic output, as reported by Brookings. A similar analysis of the 2020 presidential election once again conducted by Brookings reveals that this imbalance has persisted with the counties voting for Trump accounting for a paltry 29% of the American economy.

As an outsider constantly struggling to make sense of the wonderous disarray that is American politics, I did not expect the difference between the economic contribution of each party’s support base to be this gaping. The narrative I’d sub-consciously built in my head was one in which the GOP was the party of the rich and (under Trump) the American working and middle class. Seeing the economic contribution of the GOP’s support base dwarfed by that of the Democratic party seemed to present certain contradictions to that narrative, or at least some of the assumptions I had derived through it. In this blogpost, I thus endeavour to use data to explore the interplay between income (and more broadly class) and votes in America, and in doing so hope to distill some of the key trends in American voter behavior.

Analyzing Voter Behavior by Income Group

Using data from the American National Election Studies (ANES), I explore trends and patterns in voter behavior across different income groups. First up, we look at how voter turnout has varied across the years by income group.

Voter Turnout by Income Group

Apart from the fact that voter turnout tends to increase with income group, it is difficult to identify stand out trends from the data, particularly so amongst the middle and higher income groups. Among the lower income groups (0-16th and 17-33rd percentile) however, we do note a steady and gradual increase in voter turnout since their lows in the 1988 presidential election.

Votes by Income Group

Comparing votes across income groups, we can glean some noticeable voter trends in the longer term as well as in the more recent lead up to the 2016 Presidential Election. First, we observe that the Democratic party’s vote share among the highest income group (96th-100th percentile) has increased considerably since the 1980 presidential election. The majority vote for this income group flipped to the Democratic party in 2012 and the 2016 elections saw that trend advance.

Turning to all the other income groups however, the Democratic party’s vote share has declined since the 2008 presidential elections. The extent of this decline is greatest in the lower and middle income groups. Similar patterns in voter party identification corroborates this idea that voters political preferences across income groups are fundamentally changing, representing a shift in the political landscape of the United States.

Trump’s support base is often perceived to comprise primarily of the American working class. A core part of Trump’s 2016 campaign was after all premised on his determination to bring back jobs in the manufacturing sector and naturally this appealed greatly to the American working class. It hence comes as no surprise to see the Republican party’s vote share increasing in the lower middle bracket of income groups. Voter self-perceived class data further highlights Trump’s success in this segment of the American population.

The surge in the Republican party’s vote share among voters in lowest income group (0-16th percentile) and similarly voters who identify as lower class, is however, more intriguing. It is difficult to reconcile this trend with the fact that the Trump campaign did not directly target the lowest income groups while a key pillar of the 2016 Clinton campaign was precisely to ameliorate the socio-economic conditions of America’s poor. Examples of key campaign promises by the Democratic party included eliminating tuition at public colleges, increasing the federal minimum wage and advancing universal health care. The differences between Democratic and Republican party’s stance on tackling economic inequality is well captured by data from the Pew Research Center indicating that Democrats are more cognizant of economic inequality in the country as compared to Republicans.

The changes in vote share among voters in the 68th to 95th percentile income group is also worth analyzing. We see that the Democratic party lost some ground in this group while the Republican party held strong. Furthermore, the charts on voters’ party identification shows a considerable uptick in voters in this income group identifying as Republicans. It has always intrigued me how the GOP, which for decades was seen as the party of big business, managed to win over the majority support of America’s working class so convincingly, as has been well-documented by numerous mainstream media. What perhaps is less talked about is how America’s most destitute as well as the American upper middle class (although voting evidence pointing to the latter is weaker) seem to been swayed toward the Trump-led Republican party. The data thus seems to suggest that the Trump “working-class fan base” extends bidirectionally beyond the lower-middle income groups, calling for a rethink of our pre-conceived notions on the economic profile of Trump supporters.